摘要:
A method of and apparatus for performing blitter operations is performed in two parts. First, a description of the to-be-performed data transfer is analyzed and a computer instruction sequence is generated therefrom. The instruction sequence may include the appropriate instructions to implement scaling, clipping, and color conversion of a source image, as implicated by the blitter description. Second, the computer instruction sequence is executed an appropriate number of times, again depending upon the blitter description. One of the described embodiments transforms a blitter description that describes a target image in terms of visible rectangles into a description that describes the target image as having at least one line-type, in which the line-type comprises at least one pixel run, characterized by visible and non-visible states. The line-type description is then utilized by a code-generating logic to generate the appropriate instructions. The apparatus for performing the blitter operations may be implemented as an interface library, thus allowing client programs, including interactive games, to use the advantages of the invention to perform rapid screen updates.
摘要:
A logical device that is coupled to a computer can be controlled using string commands. The logical device requires flags and a data structure to operate properly. A string command with tokens for operating a specific logical device is provided. A linked list of the tokens is created. A command table for the specified logical device is found. The command table contains possible commands for the logical device, which commands are able to provide the required flags and data structure for the logical device. Each command in the command table contains key words or tokens or multiword tokens. The command table tokens are parsed, one at a time, against the command string tokens. If no match between the command table token and a command string token occurs, then the next command table token is parsed against the command string tokens. If a match between a command table token and a command string token occurs, then the flag that is associated with the command string token is provided and data from either the command string token or the command table token is inserted into the data structure in its proper place.
摘要:
A digitized image compression system employs several different compression techniques to compress the luminance and chrominance information and encodes the selected technique “on the fly” by inserting selected escape and header codes into the encoded data stream. More particularly, during compression, each image is divided into non-overlapping contiguous regions, referred to hereinafter as blocks and image data is compressed on a block-by-block basis. Each encoded block begins with a one-byte header that indicates, in conjunction with a previous escape code, the compression technique which was used to encode the block. Therefore, interpretation of the encoded data is a hierarchical process: the escape codes indicate a stream interpretation mode which, in turn, determine the meaning of header values which follow and the header values, in turn, determine the meaning of the encoded data in each block. The majority of image blocks are divided into quadrants and each block header is arranged as a sequence of four two-bit codes which indicate the encoding used for the four quadrants within the associated block. However, certain combinations of encoding that are unlikely to occur within the same block are inserted into the image data stream as the “escape codes” which designate the stream interpretation mode.